That would correspond with my experience with HP tanks with a 3800 ish fill that cools to 3600.
I agree the math is easier at 3600. But then again, if I have 3400 or 3500, I round down to the first number divisible by 3, which would be 3300 and turn the dive after 1100 psi is used (in this case the turn pressure is 2400 psi.) that is technically a bit less than thirds, but its on the safe side and as you indicate, violating thirds is one of the big 5 reasons people die.
You are however incorrect. People do not die because they violate thirds, they die because of poor gas planning and/or failure to turn the dive when they are supposed to.
Thirds is often not adequate, but it is often over relied on as a generic planning figure. Having bigger thirds is nice, but it does not address the real issue. Having more gas along does not reduce that risk as the risk is in the gas planning, not in the volume.
It's a lot like saying that a car with a 40 gallon tank is twice as safe as a car with a 20 gallon gas tank for a driver who runs out of gas a lot. The bigger gas tank does not solve the underlying problem that the driver forgets to check the gauge and get more gas when needed. It may reduce the frequency of running out of gas on a given route, but it does not solve the problem.
I also agree with you that tank explosion is not a big killer, but I guarentee the first time a massively overfilled LP 95 one blows up in a dive shop, a parking lot or in a cave or spring, people will die, more people will get sued and rules will get enforced. Will it happen today, next year, within the next decade? Who knows.
But it is safe bet statistically that if the cave community continues to push tanks to their full test limit with zero safety margin, eventually a tank with an inclusion or other pre or post certification defect is going to explode under that kind of no margin for error pressure.
There are also other issues and practices that come into play.
1. Many of the regs I see in use are regs that are known to have issues with high pressure o-ring pinching at pressures in excess of 3500 psi. In most cases the leaks starts out slow - just a few bubbles out the ambient chamber or a blown seal in a sealed chamber reg. But the potential for a much larger leak is there and if you dive two regs with simiolar service life and history. it is not uncommon to expect them both to fail around the same time. That screws with the idea of having reliable equipment and potentially reduces your redundancy.
2. Doubling burst discs can be exceptionally problematic. No burst disc assemblies are actually designed to be doubled so you have issues with limited thread engagement and you have issues with the disc being stressed outside the design limits as it has neither the flexibility nor the ability to be properly torqued and washered as intended. I've seen doubled burst discs blow and I'd argue they are FAR more likely to do so than a properly rated single disc. If you have to have 4000 psi pressure, use a burst disc designed for 4350 psi tanks - they are available and are much safer than doubling the discs.
3. Cave divers are exceptionally safe and many will have a fit over a thing like a 2" fastex quick release buckle. But in stark contrast, the community as a whole seems to take a total departure from common sense on the issue of tank fills in the quest for a little more gas to allow a bit more penetration. As a result they invite a whole cascade of potential failure issues (tank, burst disc, regulator, HP hoses, etc) that are 100% preventable as they have the option of just using a larger set of doubles and/or stage bottles.
I can't quite decide if number 3 is due to a collective ego thing, huberis, ignorance, or the fact that it has not yet shown up in accident analysis as a leading cause of death in cave diving - although I'd love to do statistical research on reg and burst disc failures to see if their is a correlation between fill pressure and risk of failure.
Either way, beyond certain limits that I personally think we have already passed at about 3500 psi, the safest way to address the need for more gas volume is to use larger tanks and/or stage bottles rather than to pump tanks to their test limits. Some day that practice is literally going to blow up in someone's face. If I were sure it would just be the offending diver who got killed or mained I'd be ok with that - kinda like I am ok with base jumping where the only potential victim is the guy who chose to jump. But the odds are that it will be an innocent bystander and potentially even a child who ends up injured or killed and I have no tolerance for that.
I also don't see the need. For any LP tank you can find a properly rated HP steel tank that will carry nearly the same volume at a 3500-3800 psi pressure in a similar sized package. Other than already owning one and other than the whole cave community thinking it is ok to overfill LP tanks to 3500-4000 psi because everyone does it and no one has died (yet) there is just not much reason to go the LP route, especially if you are buying a new tank. They may be slightly cheaper, but given the costs of cave diving the extra $50 for an HP tank tank is minimal
If anyone can enlighten me with a solid argument, other than the seriously flawed and rather ignorant "it has not happened - yet" argument, why any cave diver should be ok with a fill to a tank's test limit, despite the potential risks and consequences, I am all ears.
